Examining population structure across multiple collections of Cannabis

Abstract

Population structure of Cannabis sativa L. was explored across nine independent collections that each contained a unique sampling of varieties. Hierarchical Clustering of Principal Components (HCPC) identified a range of three to seven genetic clusters across datasets with inconsistent structure based on use type indicating the importance of sampling particularly when there is limited passport data. There was broader genetic diversity in modern cultivars relative to landraces. Further, in a subset of geo-referenced landrace accessions, population structure was observed based on geography. The inconsistent structure across different collections shows the complexity within Cannabis, and the importance of understanding any particular collection which could then be leveraged in breeding programs for future crop improvement.